Bipolar disorder (BPD) is one of the most common major mental illnesses, with international lifetime prevalence above 1% of the general population, affecting women and men of all ethnic groups. BPD has potential for high rates of recurrence, sustained morbidity, comorbidity (especially with substance-use and anxiety disorders), disability, and premature mortality-not only from extraordinarily high suicide rates, but also from stress-sensitive medical illnesses. To develop and implement sound and evidence-based clinical and public policies for treating BPD patients, much more information is required concerning their course and morbidity patterns, particularly from illness-onset. Information is emerging about morbidity, disability, and mortality in mid-course, but studies starting at the beginning of BPD illness remain rare. The McLean- Harvard First-Episode Project was organized by co-PI Dr. M. Tohen more than a decade ago, and continues in collaboration with PI, Professor RJ Baldessarini. The investigators have years of experience in psychiatric epidemiology and psychopharmacology research. The Project has followed the course of a large cohort of ca. 200 first-episode DSM-IV BPD (and other) patients systematically and prospectively for several years from a first hospitalization for a manic or mixed (manic-depressive) episode. The proposed study will analyze available project data to test the following important hypotheses: [1] administratively-based shortening of initial hospitalization in the 1990s will be associated with diminishing clinical improvement at discharge and rising early rehospitalization; [2] the major form of morbidity during follow-up will be depressive-dysphoric rather than manic or psychotic, with randomly distributed risk-by-time, and depression will be associated with poorer functional outcomes. Anticipated findings should contribute important new information to guide effective interventions and allocation of increasingly limited treatment resources for BPD patients.